A curated list of Python tools, libraries, and resources for audio and music processing, analysis, synthesis, and playback.
- audioread: Cross-library audio decoding (GStreamer + Core Audio + MAD + FFmpeg)
- audiomentations: Audio data augmentation library for machine learning
- babycat: Audio manipulation library for Rust Python WebAssembly and C
- matchering: Open source audio matching and mastering
- Matchering-cli: Command line application for Matchering 2.0
- noisereduce: Noise reduction using spectral gating
- numpy & scipy.io.wavfile: Read/write and manipulate WAV files
- pedalboard: Spotify's library for audio effects and processing
- PyDub: Manipulate audio with a simple and easy high level interface
- pyAudioProcessing: Audio feature extraction classification and segmentation
- SoundDevice: Play and record audio with Python
- soundfile: Read and write sound files using libsndfile
- torch-audiomentations: Fast GPU audio data augmentation for PyTorch
- torchaudio: Audio data manipulation and transformation powered by PyTorch
- wave: Read and write WAV files (Python standard library)
- aubio: Library for audio and music analysis including pitch and beat detection
- audioFlux: Library for audio and music analysis and feature extraction
- Essentia: C++ library with Python bindings for audio analysis and MIR
- librosa: Python package for music and audio analysis
- Madmom: Audio signal processing library focused on MIR tasks
- mir_eval: Evaluation functions for MIR and audio signal processing algorithms
- mirdata: Python library for working with MIR datasets
- nnAudio: GPU audio processing using PyTorch neural networks
- pyAudioAnalysis: Audio feature extraction classification segmentation and visualization
- Pyo: Python DSP module with synthesis and analysis capabilities
- scipy.signal: Signal processing routines for SciPy
- timeside: Framework for audio analysis imaging transcoding and streaming
- CLAP (LAION): Contrastive Language-Audio Pretraining for zero-shot audio classification
- CLAP (Microsoft): Learning audio concepts from natural language supervision
- OpenL3: Open-source deep audio and image embeddings
- panns-inference: Pretrained audio neural networks for audio tagging and sound event detection
- wav2vec2: Self-supervised speech representations from Facebook AI
- Whisper: OpenAI's robust multilingual speech recognition model
- faster-whisper: CTranslate2 reimplementation of Whisper up to 4x faster
- WhisperX: Whisper with word-level timestamps and speaker diarization
- SpeechRecognition: Library for performing speech recognition with multiple backends
- Vosk: Offline speech recognition API supporting 20+ languages
- SpeechBrain: PyTorch toolkit for speech processing and conversational AI
- pyannote-audio: Neural speaker diarization and voice activity detection
- Coqui TTS: Deep learning toolkit for Text-to-Speech
- Bark: Transformer-based text-to-audio model with emotions and non-speech sounds
- pyttsx3: Offline text-to-speech conversion library
- Demucs: State-of-the-art music source separation from Meta
- audio-separator: Easy stem separation using MDX-Net VR Arch and Demucs models
- Asteroid: PyTorch-based audio source separation toolkit for researchers
- pydsm: Google's toolkit for sound separation using deep learning
- Spleeter: Deezer source separation library (note: Demucs now preferred)
- basic-pitch: Spotify's lightweight neural network for polyphonic pitch detection
- CREPE: Monophonic pitch tracker using deep convolutional neural network
- torchcrepe: PyTorch implementation of CREPE pitch tracker
- MT3: Multi-instrument automatic music transcription from Google Magenta
- piano_transcription_inference: High-resolution piano transcription with pedal detection
- AudioCraft: Meta's library for MusicGen AudioGen EnCodec and MAGNeT models
- Stable Audio Tools: Generative models for conditional audio generation from Stability AI
- Riffusion: Real-time music generation using stable diffusion on spectrograms
- Magenta: Google's machine learning for music and art generation
- musicautobot: Music generation with transformers using fastai
- NSynth: Neural audio synthesis model from Magenta
- ctcsound: Python bindings for Csound using ctypes
- Mido: MIDI objects for Python
- Pippi: Computer music composition library
- pyfluidsynth: Python bindings for FluidSynth software synthesizer
- Python-audio: Jupyter notebooks about audio signal processing with Python
- Renardo: Maintained fork of FoxDot for Python live coding music
- sc3nb: SuperCollider integration for Python and Jupyter notebooks
- Abjad: Python API for building LilyPond music notation files
- AthenaCL: Algorithmic composition tool (Python 3 fork)
- maelzel: Framework for computer music in Python
- MIDIUtil: Pure Python library for creating multi-track MIDI files
- mingus: Advanced music theory and notation package
- music21: Toolkit for computer-aided musical analysis
- MusPy: Toolkit for symbolic music generation
- pretty-midi: MIDI data handling and manipulation library
- pychord: Handle and transform musical chords
- scamp: Suite for Computer-Assisted Music in Python
- audiostream: Audio API for streaming raw data to speakers
- beets: Music library manager and MusicBrainz tagger
- discord.py: Python wrapper for Discord API with music streaming
- freesound-python: Freesound API wrapper for audio retrieval and analysis
- miniaudio: Python bindings for miniaudio audio playback library
- Mopidy: Extensible music server written in Python
- Mopidy-YouTube: Mopidy extension for playing music from YouTube
- mpv: Python interface to MPV media player
- MusicBot: Discord music bot written in Python
- pyAV: Pythonic bindings for FFmpeg libraries
- pygame.mixer: Pygame module for sound loading and playback
- pyglet: Cross-platform windowing and multimedia library
- pyradio: Command line internet radio player
- Spotipy: Python client for the Spotify Web API
- AudioSet: Large-scale dataset of manually annotated audio events
- Birdsong: Dataset of annotated bird songs and calls
- Common Voice: Mozilla's open source multilingual speech dataset
- ESC-50: Environmental sound classification dataset
- Free Spoken Digit Dataset: Dataset of spoken digits in English
- Freesound Dataset: Collaborative dataset of audio samples from Freesound
- LibriSpeech: Large corpus of read English speech for ASR research
- RAVDESS: Audio-visual dataset of emotional speech and song
- Speech Commands: Dataset for speech command recognition
- TIDIGITS: Spoken digit dataset for speech recognition
- UrbanSound8K: 8000 urban sound samples in 10 classes
- VCTK: Multispeaker speech dataset for voice technologies
- VoxCeleb: Large-scale speaker identification dataset
- Beatport EDM Key: Electronic dance music tracks with musical key labels
- DALI: Dataset of lyrics and audio with time alignments
- DEAM: MediaEval dataset for music emotion recognition
- FMA: Free Music Archive dataset for music analysis
- GiantMIDI-Piano: Large-scale MIDI dataset of classical piano music
- hsmusic: Huge symbolic music dataset
- IRMAS: Instrument recognition in musical audio signals
- Jamendo Audio Tagging: Multi-label audio tagging dataset
- LAION-Audio-630K: Large collection of audio-text pairs for CLAP training
- MAESTRO: MIDI and audio dataset for music transcription and generation
- MagnaTagATune: Dataset for music annotation and audio tagging
- MedleyDB: Dataset for multi-track mixing research
- Musdb18: Dataset for music source separation
- MusicCaps: Dataset of music clips with rich text descriptions
- MusicNet: Dataset of classical music with instrument labels
- NSynth: Large-scale dataset of annotated musical notes
- Open MIC: Open Music Instrument Classification dataset
- RWC Music Database: Musical instrument sound genre and rhythm databases
- symbolic-music-datasets: Collection of symbolic music datasets
- The Million Song Dataset: Massive collection of audio features and metadata
- librosa tutorial - Introduction: Advanced librosa tutorial covering spectrograms and remixing
- librosa tutorial - Visualization: Visualizing sounds using librosa and matplotlib
- PyDub tutorial: Working with WAV files using PyDub
- Whisper tutorial: Using OpenAI Whisper for speech-to-text
- AudioCraft tutorial: Getting started with MusicGen and AudioGen
- Hugging Face Audio Course: Comprehensive course on audio ML with transformers